东海和黄海南部近岸海域小黄鱼稚幼鱼耳石元素指纹特征分析

Analysis on the otolith core elemental fingerprint of larval and juvenile Larimichthys polyactis in the coastal areas of East China Sea and southern Yellow Sea

采用LA-ICP-MS分析技术对东海和黄海南部近岸五个海域(芦潮港、象山、三门湾、温岭、吕四)小黄鱼(Larimichthys polyactis)稚幼鱼耳石核区元素指纹进行了测定。运用方差分析方法(MANOVA和ANOVA)检验不同海域小黄鱼群体耳石元素比值的差异显著性,采用线性判别分析(LDFA)方法对不同海域的小黄鱼群体进行了群体属性判别,并运用留一交叉验证法(LOOCV)评估了模型的判别成功率。MANOVA分析结果显示,不同采样点小黄鱼耳石核区的元素比值呈极显著差异(Wilk’s Lambda检验,F4,131=28.00,P<0.01)。各元素比值的ANOVA分析结果显示,除了Zn∶Ca呈显著性差异外(P<0.05),Li∶Ca、Mg∶Ca、Mn∶Ca、Sr∶Ca、Na∶Ca和Ba∶Ca呈极显著差异(P<0.01)。判别分析结果显示,小黄鱼群体的总体判别成功率为70.6%,其中吕四小黄鱼群体的判别成功率最高,为86.2%;温岭最低,为58.3%。黄海南部的吕四小黄鱼群体可以与东海几个群体区分开来,而东海的几个小黄鱼群体难以有效区分。本研究的结果表明基于LA-ICP-MS的耳石元素指纹分析技术能够对不同的鱼类群体进行识别研究,具有广泛的应用前景。

In this paper, we studied the otolith core elemental fingerprint of larval and juvenile small yellow croaker （Larimichthys polyactis） sampled in different sites from the East China Sea and southern Yellow Sea coastal areas using the LA-ICP-MS element analysis technique. The specimens were collected from sites：Lvsi, Luchao Port, Xiangshan, Wenling and Sanmen Bay. MANOVA and ANOVA analysis were used to test the difference of elemental ratios between the populations of small yellow croaker sampled in different areas. Linear discriminant analysis （LDFA） was used to discriminate the populations of small yellow croaker. Besides, we used the “leave one out cross validation （LOOCV）” method to evaluate the success rate of the discriminant analysis model. The results of MANOVA analysis showed that there were extremely significant differences （Wilk＇ s Lambda test, F4.131 = 28.0, P 〈 0.01 ） of elemental ratios between the populations of small yellow croaker sampled in different areas. The results of ANOVA analysis showed that Zn： Ca showed significant differences （ P 〈 0.05 ） while Li ： Ca, Mg ： Ca, Mn ： Ca, Sr ： Ca, Na ： Ca and Ba ： Ca showed extremely significant differences （P 〈 0.01 ） between the populations of small yellow croaker. Via linear discriminant function analysis （LDFA） based on elemental ratios （Li： Ca, Na： Ca, Mg： Ca, Mn： Ca, Sr： Ca, Ba： Ca）, fishes were classified back to sampling sites with 70.6% correction. The classification accuracy of Lvsi was the highest （ 86.2% ）, while Wenling was the lowest （ 58. 3% ）. The Lvsi population in the southern Yellow Sea could obviously be distinguished from other populations, other populations in the East China Sea were mixed er and it was difficult to distinguish therm. The results of this study shows that we can use the otolith elemental fingerprint based on LA-ICP-MS analysis technique to identify different populations of fish and this method will have wide application prospect.